6 {{ upvoteCount | shortNum }}
6 {{ upvoteCount | shortNum }}

Split Testing - Minimum Unique Visits?

by GoodnightSweetRatRace
5 replies
When split testing, what would you say is the minimum amount of Unique Visits you should recieve before trying something else?

I want to try out some different headers on my squeeze page to try and increase the opt-in rate, but not sure about how many UV's I should recieve before trying a new header.
#minimum #split #unique #visits
  • Profile picture of the author David Keith
    obviously, more is better. but unless you are seeing a consistent 50+ a day or so and not having any luck i would say your conversion numbers may or may not be very statistically significant.

    some will tell you more and some less. and it also depends on the conversion rate you are seeing.

    thus in my 50 a day example, you would get 1500 visitors in a month. if you dont convert somewhere near 1% for most industries, i would think it would be time to re-evaluate.

    but there is no secret number to have. and traffic source will have more of an impact than any on site metric you could change, so if your not converting as you had hoped. you traffic source is the first thing to check.
    Signature
    Just a Little War Room Gift.
    {{ DiscussionBoard.errors[4573828].message }}
    • Profile picture of the author GoodnightSweetRatRace
      Originally Posted by owslaw123 View Post

      obviously, more is better. but unless you are seeing a consistent 50+ a day or so and not having any luck i would say your conversion numbers may or may not be very statistically significant.

      some will tell you more and some less. and it also depends on the conversion rate you are seeing.

      thus in my 50 a day example, you would get 1500 visitors in a month. if you dont convert somewhere near 1% for most industries, i would think it would be time to re-evaluate.

      but there is no secret number to have. and traffic source will have more of an impact than any on site metric you could change, so if your not converting as you had hoped. you traffic source is the first thing to check.

      I'm trying to increase my opt-in rate on my squeeze page. It currently converts at around 35% and I think with a bit of tweaking I can get it to 40%+.
      Signature

      {{ DiscussionBoard.errors[4573863].message }}
      • Profile picture of the author David Keith
        Originally Posted by GoodnightSweetRatRace View Post

        I'm trying to increase my opt-in rate on my squeeze page. It currently converts at around 35% and I think with a bit of tweaking I can get it to 40%+.
        ah. i see. yes of course that sounds very likely possible.

        obviously testing is the key. and you might look up multivariate testing.

        this is a more advanced testing methods than normal split testing. they test multiple different page elements at a time and eventually pick the best converting page.

        its is more complicated than normal split testing, but if you are already having good success, then these more advanced testing methods will probably benefit you.
        Signature
        Just a Little War Room Gift.
        {{ DiscussionBoard.errors[4573916].message }}
      • Profile picture of the author JohnMcCabe
        Instead of looking at the number of visitors, look at the number of results (opt-ins, in this case). Let your test run until you get 100 opt-ins, minimum.

        And to be a real test, you have to be running the test header against your original header, alternating visitors between them. Just running one header for 100 results, then switching to a new one for 100 results introduces variables you can't control and which can skew your results. You do not want to be basing your decisions on flawed results...
        {{ DiscussionBoard.errors[4573919].message }}
        • Profile picture of the author Alex Makarski
          Run it through www.SplitTester.com.com. It's more geared towards Adwords but there are ways to plug in your numbers in there an make it work. It'll tell you if you have received enough clicks to pick a winner based on the difference in the performance of the two versions of your page.
          Signature

          "I hear and I forget. I see and I remember. I do and I understand." -Confucius

          {{ DiscussionBoard.errors[4574033].message }}

Trending Topics